Oil chromatogram online monitoring data correction method and system based on multi-branch physical constraint neural operator
By using a multi-branch physical constraint neural operator model to refine the chromatographic signal of transformer oil, the problems of chromatographic peak shape distortion and unreliable quantitative information in the existing technology are solved, and efficient and reliable online monitoring data output is achieved to support fault diagnosis.
CN122409936APending Publication Date: 2026-07-17WUHAN NARI LIABILITY OF STATE GRID ELECTRIC POWER RES INST
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- WUHAN NARI LIABILITY OF STATE GRID ELECTRIC POWER RES INST
- Filing Date
- 2026-05-18
- Publication Date
- 2026-07-17
Smart Images
Figure CN122409936A_ABST
Abstract
本发明提出一种基于多分支物理约束神经算子的油色谱在线监测数据校正方法。该方法包括获取在线监测装置输出的实时原始色谱曲线，将实时原始色谱曲线分割为K个子曲线，构建多分支物理约束神经算子模型，并使用历史原始色谱曲线进行训练得到训练后的多分支物理约束神经算子模型，将实时原始色谱子曲线输入至训练好的多分支物理约束神经算子模型于校正后色谱子曲线，并融化输出完整的多组分校正色谱信号。本发明能对多组分色谱信号进行按组分独立、精细化的修复，在校正过程中强制遵循色谱物理规律，对乙炔等高故障敏感度气体实施差异化的、高优先级的定量保真。
Need to check novelty before this filing date? Find Prior Art